A Complete Guide to the Machine Learning Tools on AWS
A Complete Guide to the Machine Learning Tools on AWS

AWS Machine Learning

A Complete Guide to the Machine Learning Tools on AWS

In this article, we will take a look at each one of the machine learning tools offered by AWS and understand the type of problems they try to solve for their customers.

Manish Manalath
Jan 17 · 6 min read

AWS Machine Learning comprises of a rich set of tools that Amazon offers to help developers integrate machine learning models into their applications. AWS offers pre-trained models for use cases including computer vision, recommendation engines, and language translation.

Amazon currently offers 15 machine learning services on its platform. In this article, we will take a look at each one of them and understand the type of problem they try to solve for their customers.


SageMaker also has an autopilot option that will automatically process the data and run it through multiple algorithms. This helps developers to find the best algorithm for their model without manually training and testing their models. Sagemaker also comes with an integrated IDE and a sharable jupyter notebook that you can use to collaborate with your team.


Codeguru’s algorithms are trained with codebases from Amazon’s projects. Right now, CodeGuru supports only Java applications, but we can expect the functionality extended to other languages in the near future.


Comprehend is also a fully managed service, meaning you can use pre-trained models to work with your data. Comprehend also has an additional service called Amazon Comprehend Medical that lets you work with medical documents to analyze medical conditions and dosages.


Forecast is also customizable and lets you build custom models on top of Amazon’s existing deep learning models. Like most machine learning tools in AWS, Forecast is also fully managed and can scale according to your business needs.

Fraud Detector

A fraud detector needs an existing dataset of labeled fraudulent transactions in order to train and understand the pattern of your customer behavior. This is used to prevent further fraudulent transactions. You can also configure custom authentication rules for guest logins and product trials.


Kendra can be used to help customers find answers to specific problems while using your product without the need for additional customer support. Kendra also supports natural language questions, delivering an even smoother experience for your customers.


Lex can be used as a replacement for manual customer support to help filter the usual queries, answering them automatically. Lex is also a fully managed service that scales automatically and employs a pay-as-you-use model.


Personalize is a great tool to build product recommendations, custom search results based on queries and employ targeted marketing promotions.


Polly is powered by deep learning algorithms that mimic a conversational style interface that can be used in narrations, telephony applications, etc.


Rekognition can be employed in use cases like identifying manufacturing defects in products, spotting unauthorized personnel in an organization, scanning for inappropriate content in movies, etc. Rekognition can also be used to analyze player movements in games for post-game analysis.


Textract is useful for processing loan applications, medical claims, etc. In addition to extracting data, they can be optimized for search using Textract. Documents that usually take months to process using manual methods can be processed in hours using AWS Textract.


Transcribe can also be used to convert customer calls into text and analyze them for improved customer service. Cataloging audio archives is another use case for AWS transcribe.


Translate is also designed to be more natural-sounding to customers since the context of the sentence is also taken into account. Translate is also highly customizable to help improve the accuracy of translation when working with your brand names and unique words related to your business.


DeepLens is designed for developers getting started in machine learning to get a grasp of how their models will work in the real world. DeepLens is also integrated with the AWS ecosystem and can be used with other AWS services like Lambda and Rekognition to extend its capabilities.


You can build reinforcement learning models using AWS SageMaker and test them instantly using DeepRacer. Amazon also offers an opportunity to connect and compete with fellow enthusiasts by building virtual private race tracks.


Hope you enjoyed the article. If you have any questions, let me know in the comments. You can also signup for my newsletter to receive a summary of articles once a week.

AWS Tutor

Articles, How-Tos and Use Cases on AWS Tools and Services

Manish Manalath

Written by

10 years as a Digital Nomad. Writes on Tech (http://awstutor.io) and Productivity. Connect with me on Linkedin (linkedin.com/in/manishmshiva)

AWS Tutor

AWS Tutor

Articles, How-Tos and Use Cases on AWS Tools and Services

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade